import pandas as pd
import urllib
import numpy as np
import urllib.request
import re
from textblob import TextBlob
%run lib.py

#name="Legally%20Blonde"
#name="aboutmary"
#name="10Things"
name="magnolia"
#name="Friday%20The%2013th"
#name="Ghost%20Ship"
#name="Juno"
#name="Reservoir+Dogs"
#name="shawshank"
#name="Sixth%20Sense,%20The"
#name="sunset_bld_3_21_49"
#name="Titanic"
#name="toy_story"
#name="trainspotting"
#name="transformers"
#name="the-truman-show_shooting"
#name="batman_production"

ext="html"
txtfiles=["Ghost%20Ship", "Legally%20Blonde", "Friday%20The%2013th", "Juno", "Reservoir+Dogs", "Sixth%20Sense,%20The", "Titanic"]
if name in txtfiles:
    ext="txt"
fp = urllib.request.urlopen("http://www.dailyscript.com/scripts/"+name+"."+ext)
mybytes = fp.read()

mystr = mybytes.decode("utf8", "ignore")
fp.close()
liston=mystr.split("\n")
liston=[s.replace('\r', '') for s in liston]
liston=[re.sub('<[^<]+?>', '', text) for text in liston]

if name=="shawshank":
    liston=[i.replace("\t", "    ") for i in liston]

char=""
script=[]
charintro='                                 '
endofdialogue='          '
dialoguepre='                    '
newscenepre='          '
charintro=''
endofdialogue=''
dialoguepre=''
newscenepre=''
i=45
print("Characters")
i, charintro=nextbigchunk(liston, i)
print("Adverbs")
i, adverb=nextbigchunk(liston, i, adverbs=True)
print("Dialogues")
i, dialoguepre=nextbigchunk(liston, i)
print("New Scene:")
i, newscenepre=nextbigchunk(liston, i)

if newscenepre=="X":
    i=100
    i, newscenepre=nextbigchunk(liston, i)
    if name=="aboutmary":
        newscenepre=" ".join(["" for i in range(56)])
    if len(newscenepre)==len(charintro):
        newscenepre="X"
    

endofdialogue=newscenepre
    

scene=1
for s in liston:
    if s[0:len(charintro)]==charintro and s[len(charintro)]!=" " and s.strip()[0]!="(" and s.strip()[len(s.strip())-1]!=")":
        #print("Charatcer*****")
        char=s[len(charintro):]
        new=dict()
        new['char']=char.strip()
        new['dialogue']=""
        new['scene']=scene
        new['adverb']=""
    if s==endofdialogue or s.replace(" ", "")=="":
        if char!="":
            char=""
            script.append(new)
    if char!="" and s[0:len(dialoguepre)]==dialoguepre and s[len(dialoguepre)]!=" ":
        #print("Dialogue******")
        if new['dialogue']!="":
            new['dialogue']=new['dialogue']+" "
        new['dialogue']=new['dialogue']+s[len(dialoguepre):]
    if char!="" and ((s[0:len(adverb)]==adverb and s[len(adverb)]!=" ") or (len(s)>1 and s.strip()[0]=="(" and s.strip()[len(s.strip())-1]==")" )):
        if new['adverb']!="":
            new['adverb']=new['adverb']+" "
        new['adverb']=new['adverb']+s[len(adverb):]
    if s[0:len(newscenepre)]==newscenepre and len(s)>len(newscenepre) and ( s.isupper()) and s[len(newscenepre)]!=" ":
        scene=scene+1

Characters
                                magnolia
                                NARRATOR
                                NARRATOR
                                NARRATOR
                                NARRATOR
                                NARRATOR
Adverbs
Dialogues
                      In the New York Herald, November 26,
                      year 1911, there is an account of the
                      hanging of three men --
                      ...they died for the murder of
                      Sir Edmund William Godfrey --
                      -- Husband, Father, Pharmacist and all
New Scene:
     a P.T. Anderson picture                             11/10/98
     a Joanne Sellar/Ghoulardi Film Company production

pd.DataFrame(script).to_csv(name+'.csv', index=None)
pd.DataFrame(script)

magnolia=pd.read_csv(name+'.csv')
stopwords = getstopwords()

removedchars=["'S VOICE", "'S WHISPER VOICE", " GATOR"]
for s in removedchars:
    magnolia['char']=magnolia['char'].apply(lambda x: x.replace(s, ""))
i=0
scenes=dict()
for s in magnolia.iterrows():
    scenes[s[1]['scene']]=[]
for s in magnolia.iterrows():
    scenes[s[1]['scene']].append(s[1]['char'])
for s in magnolia.iterrows():
    scenes[s[1]['scene']]=list(set(scenes[s[1]['scene']]))

characters=[]
for s in scenes:
    for k in scenes[s]:
        characters.append(k)
characters=list(set(characters))
appearances=dict()
for s in characters:
    appearances[s]=0
for s in magnolia.iterrows():
    appearances[s[1]['char']]=appearances[s[1]['char']]+1

a=pd.DataFrame(appearances, index=[i for i in range(len(appearances))])

finalcharacters=[]
for s in pd.DataFrame(a.transpose()[0].sort_values(0, ascending=False))[0:10].iterrows():
    finalcharacters.append(s[0])

finalcharacters
file=open(name+"_nodes.csv", "w")
couplesappearances=dict()
for s in finalcharacters:
    file.write(";")
    file.write(s)
file.write("\n")
for s in finalcharacters:
    newlist=[]
    for f in finalcharacters:
        newlist.append(0)
        couplesappearances[f+"_"+s]=0
    j=0
    for f in finalcharacters:
        for p in scenes:
            if f in scenes[p] and s in scenes[p] and f!=s and finalcharacters.index(f)<finalcharacters.index(s): 
                long=len(magnolia[magnolia["scene"]==p])
                newlist[j]=newlist[j]+long
                couplesappearances[f+"_"+s]=couplesappearances[f+"_"+s]+long
        j=j+1
    file.write(s)
    for f in newlist:
        file.write(";")
        file.write(str(f))
    file.write("\n")
file.close()

a=pd.DataFrame(couplesappearances, index=[i for i in range(len(couplesappearances))])
finalcouples=[]
for s in pd.DataFrame(a.transpose()[0].sort_values(0, ascending=False))[0:4].iterrows():
    finalcouples.append(s[0])

file=open(name+"_finalcharacters.csv", "w")
for s in finalcharacters:
    file.write(s+"\n")
file.close()
file=open(name+"_finalcouples.csv", "w")
for s in finalcouples:
    file.write(s+"\n")
file.close()

importantchars=[]
for char in appearances:
    if appearances[char]>10:
        importantchars.append(char)

file=open(name+"_sentiment_overtime_individual.csv", "w")
file2=open(name+"_sentiment_overtime_individualminsmaxs.csv", "w")

for k in finalcharacters:
    print(k)
    dd=getdialogue(magnolia, k, k, scenes)
    dd=[str(d) for d in dd]
    polarities, subjectivities=getsentiment(dd)
    %matplotlib inline
    import matplotlib.pyplot as plt
    moveda=maverage(polarities, dd, .99)
    plt.plot(moveda)
    i=0
    for s in moveda:
        file.write(k+","+str(float(i)/len(moveda))+", "+str(s)+"\n")
        i=i+1
    plt.ylabel('polarities')
    plt.show()
    file2.write(k+"| MIN| "+dd[moveda.index(np.min(moveda))]+"\n")
    file2.write(k+"| MAX| "+dd[moveda.index(np.max(moveda))]+"\n")
    print("MIN: "+dd[moveda.index(np.min(moveda))])
    print("\n")
    print("MAX: "+dd[moveda.index(np.max(moveda))])
    
file.close()
file2.close()

file=open(name+"_sentiment_overtime_couples.csv", "w")
file2=open(name+"_sentiment_overtime_couplesminsmaxs.csv", "w")

for k in finalcouples:
    print(k)
    liston=k.split("_")
    dd=getdialogue(magnolia, liston[0], liston[1], scenes)
    dd=[str(d) for d in dd]
    polarities, subjectivities=getsentiment(dd)
    %matplotlib inline
    import matplotlib.pyplot as plt
    moveda=maverage(polarities, dd, .99)
    plt.plot(moveda)
    i=0
    for s in moveda:
        file.write(k+","+str(float(i)/len(moveda))+", "+str(s)+"\n")
        i=i+1
    plt.ylabel('polarities')
    plt.show()
    file2.write(k+"| MIN| "+dd[moveda.index(np.min(moveda))]+"\n")
    file2.write(k+"| MAX| "+dd[moveda.index(np.max(moveda))]+"\n")
    print("MIN: "+dd[moveda.index(np.min(moveda))])
    print("\n")
    print("MAX: "+dd[moveda.index(np.max(moveda))])
    
file.close()
file2.close()

JIM KURRING

MIN: You mind if I check things back here? 


MAX: YOU'RE OK...you're gonna be ok....
JIMMY

MIN: She went crazy.  She went crazy, Rose. 


MAX: Imagine you are attending a jam session of classical composers and they have  each done an arrangment of the classic  favorite, "Whispering."  Here are three  variations on the theme, as three classic  composer's might have written it -- you are to name the composer.  The First: 
CLAUDIA

MIN: I'm sorry. 


MAX: Did you ever go out with someone and just....lie....question after question, maybe you're trying to  make yourself look cool or better  than you are or whatever, or smarter  or cooler and you just -- not really lie, but maybe you just don't say everything --
FRANK

MIN: If you feel, made to feel like you need them, like -- like you can't live if you're without them or you need, what?  They're pussy?  They're love? Fuck that.  Self Sufficient, gents.  That's the truth. What you are -- we are -- you need them  for what?  To fucking make you a piece of snot rag?  A puppett?  huh?  Hear them bitch and moan? bitch and moan --  and we're taught one thing -- go the other way -- there is No Excuse I will give you, I'm not gonna apologize -- I'm not gonna  apologize for my NEED my DESIRE...my, the  things that I need as a man to feel comfortable... You understand?  You understand?  You need to say something, "my mommy hit me or  daddy hit me or didn't let me play soccer,  so now I make mistakes, cause a that -- something, so now I piss and shit on it and do this." Bullshit.  I'm sorry. ok. yeah. no. fuck.  go.  fuck. alright. go make a new mistake. maybe not, I dunno...fuck.... 


MAX: I wouldn't want that to be misunderstood: My enrollment was totally unoffical because I was, sadly, unable to afford tuition up  there.  But there were three wonderful men who were kind enough to let me sit in on their classes, and they're names are:  Macready, Horn and Langtree among others. I was completely independent financially, and like I said: One Sad Sack A Shit.  So what we're looking at here is a true rags to riches story and I think that's  what most people respond to in "Seduce," And At The End Of The Day? Hey -- it may not  even be about picking up chicks and sticking your cock in it -- it's about finding What You Can Be In This World.  Defining It.  Controling It and  saying: I will take what is mine.  You just happen  to get a blow job out of it, then hey-what-the-fuck- why-not?  he.he.he.
PHIL

MIN: You wanna call him on the phone? We can call him, I can dial the  phone if you can remember the number -- 


MAX: Thank you, Chad, and good luck to you and your mother -- 
STANLEY

MIN: I think that you have to be nicer to me.


MAX: I'm fine. I'm fine, I just wanna keep playing --
DONNIE

MIN: My teeff...my teeef....


MAX: My name is Donnie Smith and I have lot's of love to give. 
EARL

MIN: No, no, the grade...the grade that you're in? 


MAX: "...it's not going to stop 'till you wise up..."
LINDA

MIN: listen...listen to me now, Phil:  I'm sorry, sorry I slapped your face.  ...because I don't know what I'm doing... ...I don't know how to do this, y'know?  You understand?  y'know?  I...I'm...I do things  and I fuck up and I fucked up....forgive me, ok? Can you just...


MAX: I'm listening.  I'm getting better. 
NARRATOR

for key, val in scenes.items():
    for s in scenes[key]:
        new="INSCENE_"+scenes[key][0]
        scenes[key].remove(scenes[key][0])
        scenes[key].append(new)

magnolia.dropna(subset=['dialogue'])
1

1

baskets=[]
spchars=["\"", "'", ".", ",", "-"]
attributes=["?", "!"]
for s in magnolia.iterrows():
    if type(s[1]['dialogue'])!=float and  len(s[1]['dialogue'])>0:
        new=[]
        for k in scenes[s[1]['scene']]:
            new.append(k)
        new.append("SPEAKING_"+s[1]['char'])
        for k in s[1]['dialogue'].split(" "):
            ko=k
            for t in spchars:
                ko=ko.replace(t, "")
            for t in attributes:
                if ko.find(t)>=0:
                    new.append(t)
                    ko=ko.replace(t, "")
            if len(ko)>0:
                new.append(ko.lower())
        new=list(set(new))
        baskets.append(new)

baskets2=[]
basketslist=[]
for k in baskets:
    new=dict()
    new2=[]
    for t in k:
        if t not in stopwords:
            new[t]=1
            new2.append(t)
    baskets2.append(new)
    basketslist.append(new2)

baskets2=pd.DataFrame(baskets2)
from mlxtend.frequent_patterns import apriori
from mlxtend.frequent_patterns import association_rules
baskets2=baskets2.fillna(0)
baskets2.to_csv(name+'_basket.csv')

frequent_itemsets = apriori(baskets2, min_support=5/len(baskets2), use_colnames=True)
rules = association_rules(frequent_itemsets, metric="lift", min_threshold=1)

rules['one_lower']=[int(alllower(i) or alllower(j)) for i, j in zip(rules['antecedants'], rules['consequents'])]

rules['both_lower']=[int(alllower(i) and alllower(j)) for i, j in zip(rules['antecedants'], rules['consequents'])]

rules.to_csv(name+'_rules.csv', index=None)

	adverb	char	dialogue	scene
0		magnolia		1
1		NARRATOR	In the New York Herald, November 26, year 1911...	2
2		NARRATOR	...they died for the murder of Sir Edmund Will...	2
3		NARRATOR	-- Husband, Father, Pharmacist and all around ...	2
4		NARRATOR	Greenberry Hill, London. Population as listed.	3
5		NARRATOR	He was murdered by three vagrants whose motive...	5
6		NARRATOR	...Joseph Green.....	5
7		NARRATOR	...Stanley Berry....	5
8		NARRATOR	...and Nigel Hill...	5
9		NARRATOR	Green, Berry and Hill.	7
10		NARRATOR	...And I Would Like To Think This Was Only A M...	7
11		NARRATOR	As reported in the Reno Gazzette, June of 1983...	9
12		NARRATOR	--- the water that it took to contain the fire --	10
13		NARRATOR	-- and a scuba diver named Delmer Darion.	12
14		NARRATOR	Employee of the Peppermill Hotel and Casino, R...	15
15		NARRATOR	-- well liked and well regarded as a physical,...	16
16		NARRATOR	-- as reported by the coroner, Delmer died of ...	21
17		NARRATOR	...volunteer firefighter, estranged father of ...	24
18		NARRATOR	-- added to this, Mr. Hansen's tortured life m...	26
19		CRAIG HANSEN	...oh God...fuck...I'm sorry...I'm sorry...	27
20		NARRATOR	The weight of the guilt and the measure of coi...	27
21		CRAIG HANSEN	...forgive me...	27
22		NARRATOR	And I Am Trying To Think This Was All Only A M...	29
23		NARRATOR	The tale told at a 1961 awards dinner for the ...	32
24		NARRATOR	Seventeen year old Sydney Barringer. In the ci...	33
25		NARRATOR	The coroner ruled that the unsuccessful suicid...	33
26		NARRATOR	The suicide was confirmed by a note, left in t...	34
27		NARRATOR	At the same time young Sydney stood on the le...	35
28		NARRATOR	The neighbors heard, as they usually did, the...	36
29		NARRATOR	-- and it was not uncommon for them to threat...	37
...	...	...	...	...
1493		DIXON	We gotta get his money so we can get outta her...	382
1494		WORM	That idea is over now. We're not gonna do tha...	382
1495	(to Stanley)	DIXON	DADDY, FUCK, DADDY, DON'T GET MAD AT ME. DON'T...	382
1496		WORM	I'm not mad, son, I will not be mad at you an...	382
1497		DIXON	DAD.	382
1498		DIXON	I - just - thought - that - I - didn't want - ...	382
1499		WORM	It's ok, boy.	382
1500		MUSIC/KERMIT THE FROG	"It's not that easy bein' green... Having to s...	383
1501		DONNIE	My teeff...my teeef....	385
1502		JIM KURRING	YOU'RE OK...you're gonna be ok....	385
1503		NARRATOR	And there is the account of the hanging of thr...	390
1504		NARRATOR	There are stories of coincidence and chance an...	391
1505		NARRATOR	...and we generally say, "Well if that was in...	392
1506		DOCTOR	Are you with us? Linda? Is it Linda?	394
1507		NARRATOR	Someone's so and so meet someone else's so and...	395
1508		NARRATOR	And it is in the humble opinion of this narrat...	398
1509		STANLEY	Dad...Dad.	399
1510		STANLEY	You have to be nicer to me, Dad.	399
1511		RICK	Go to bed.	399
1512		STANLEY	I think that you have to be nicer to me.	399
1513		RICK	Go to bed.	399
1514		NARRATOR	...and so it goes and so it goes and the book...	400
1515		MARCIE	I killed him. I killed my husband. He hit my...	401
1516		DONNIE	I know that I did a thtupid thing. Tho-thtupid...	402
1517		DONNIE	I really do hath love to give, I juth don't kn...	402
1518		JIM KURRING	...these security systems can be a real joke. ...	403
1519		DONNIE	....ohh-thur-I-thur-thill....	403
1520		JIM KURRING	You guys make alotta money, huh?	403
1521	(beat)	JIM KURRING	...alot of people think this is just a job tha...	405
1522		END.		406

sentiment	Porcentaje
positive	65.4%
negative	34.6%

sentiment	Porcentaje
positive	21.6%
negative	12.9%
trust	12.8%
joy	12.7%
anticipation	11.5%
fear	6.2%
disgust	5.8%
sadness	5.5%
surprise	5.5%
anger	5.4%

sentiment	Porcentaje
negative	34.1%
positive	31.4%
uncertainty	28.4%
litigious	5.7%
superfluous	0.4%

sentiment	Porcentaje
positive	64.3%
negative	35.7%

Analisis de Script Pelicula Juno

Bernardo Garcia & Eduardo Hidalgo

May 9, 2018

Agenda

Procesamiento de Texto

Ver archivo Python

Analisis de Sentimiento (Pelicula & Personaje)

Score por Pelicula

Score por Personaje

Score por Personaje en el tiempo

Top 10 Personajes

Top 4 Parejas

Reglas de Asociación entre palabras (Market Basket)

Toda la pelicula

Top 10 Personajes

Top 4 Parejas

Analisis de Relaciones entre Personajes (Pagerank)

sentiment	Porcentaje
positive	21.6%
joy	13.7%
negative	12.6%
trust	12.5%
anticipation	10.6%
disgust	6.3%
fear	6.2%
sadness	6.0%
surprise	5.6%
anger	5.1%

sentiment	Porcentaje
positive	21.0%
negative	14.6%
anticipation	13.1%
trust	13.1%
joy	9.4%
fear	7.9%
anger	6.0%
disgust	5.2%
sadness	4.9%
surprise	4.9%

sentiment	Porcentaje
positive	43.8%
negative	25.0%
uncertainty	18.8%
litigious	9.4%
superfluous	3.1%

sentiment	Porcentaje
positive	22.2%
joy	16.4%
trust	14.8%
anticipation	13.8%
negative	11.6%
sadness	5.8%
surprise	5.3%
fear	4.2%
anger	3.2%
disgust	2.6%

sentiment	Porcentaje
positive	21.4%
anticipation	16.7%
trust	11.9%
joy	11.1%
negative	11.1%
fear	7.9%
anger	5.6%
surprise	5.6%
sadness	4.8%
disgust	4.0%

sentiment	Porcentaje
positive	23.4%
joy	14.6%
trust	12.4%
negative	10.2%
anticipation	8.8%
disgust	7.3%
fear	7.3%
anger	6.6%
sadness	5.8%
surprise	3.6%

sentiment	Porcentaje
positive	20.6%
joy	15.4%
trust	15.4%
negative	13.2%
anticipation	11.8%
surprise	7.4%
anger	5.9%
disgust	4.4%
sadness	3.7%
fear	2.2%

sentiment	Porcentaje
positive	20.0%
negative	16.0%
anticipation	11.2%
trust	11.2%
joy	9.6%
disgust	8.8%
fear	6.4%
sadness	6.4%
surprise	5.6%
anger	4.8%

sentiment	Porcentaje
positive	17.4%
trust	17.4%
anticipation	13.0%
joy	13.0%
surprise	13.0%
negative	8.7%
anger	4.3%
disgust	4.3%
fear	4.3%
sadness	4.3%

Personaje	Min_Max	Dialogo
JUNO	MIN	Oh, I couldn’t copy your work.
JUNO	MAX	I’m going to say I’m 104% sure.
MARK	MIN	Nesting, huh? Are you planning to build the crib out of twigs and saliva?
MARK	MAX	I called Gerta Rauss. She says she can represent both of us. They call it “collaborative divorce.” It’s apparently all the rage right now. And it’s easy because we don’t have children.
VANESSA	MIN	Then what’s going on?
VANESSA	MAX	That’s great.
BLEEKER	MIN	Why?
BLEEKER	MAX	I always think you’re cute. I think you’re beautiful.
LEAH	MIN	Oh yeah! “Desperately Seeking Spawn.” They’re right by the ads for like, iguanas and terriers and used fitness equipment. It’s totally legit.
LEAH	MAX	You know, you can go into early labor sucking face like that!
MAC	MIN	You mean like couples?
MAC	MAX	We’re fine, thank you.
BREN	MIN	They could be utterly negligent. Maybe they’ll do a far shittier job of raising a kid than my dumbass stepdaughter ever would. Have you considered that?
BREN	MAX	Oh, go fly a kite.
ULTRASOUND TECH	MIN	Well, there we have it. Would you like to know the sex?
ULTRASOUND TECH	MAX	Planning to be surprised when you deliver?
BLEEKER’S MOM	MIN	Hi Juno. What can I do for you?
BLEEKER’S MOM	MAX	Fine. Come in.
GERTA	MIN	So, look those over and give me a call at my office if you have any questions.
GERTA	MAX	Nice to meet you.

Parejas	Min_Max	Dialogo
JUNO_MARK	MIN	I don’t get a klepto vibe from you. Evil genius? Maybe. Arsonist? Wouldn’t rule it out.
JUNO_MARK	MAX	I think it kind of looks like my friend, Paulie.
JUNO_LEAH	MIN	The Penny Saver sucks.
JUNO_LEAH	MAX	You know, you can go into early labor sucking face like that!
JUNO_VANESSA	MIN	Juno, we’d really appreciate it if you could keep us updated on any doctor’s appointments, ultrasounds, other things of that nature.
JUNO_VANESSA	MAX	That’s great.
MARK_VANESSA	MIN	Juno, we’d really appreciate it if you could keep us updated on any doctor’s appointments, ultrasounds, other things of that nature.
MARK_VANESSA	MAX	Um, I think people are kind of unsure about the situation because it’s not, you know, set in stone.

Numero de Dialogos	Leverage Minimo	Leverage Maximo
3,314	-0.0017	0.0017
7,980	0.0017	0.0051
9,850	0.0051	0.0085
2,122	0.0085	0.012
942	0.012	0.015
496	0.015	0.019